【摘要】 在分析石脑油裂解炉辐射段的Kumar分子反应动力学模型的基础上,提出了一种应用数据融合技术对Kumar模型一次反应选择性系数进行在线调整的方法。首先,利用进化的遗传算法,以改进的石脑油裂解模型计算出的产率与实际工况产率的误差作为目标函数,对选择性系数进行离线调整,进而建立原料性质与选择性系数相匹配的标准样本数据库。其次,利用数据融合技术中的Dempeter-Shafer证据理论建立模型,计算待估物料与标准样本数据库中样本物料的匹配度,进而对待估原料的一次反应选择性系数进行估算。研究结果表明,利用估算出的一次反应选择性系数建立的裂解工艺模型具有较高精度,将仿真结果与实际工况产率进行对比,证实了该方法的有效性和可行性。更多还原

【Abstract】 On the analysis of the Kumar’s kinetic model of naphtha pyrolysis furnace,on-line adjustment method based by the data fusion theory for the selectivity of the first-order reaction of the Kumar’s model is proposed.Firstly,the improved genetic algorithm is used to realize the off-line adjustment of the selectivity of the first-order reaction,and the sample database is constituted by the property of stuff and the selectivity depending on the property of stuff.Secondly,an on-line method for estimating of the selectivity of the first-order reaction for the new property of stuff is presented by the Dempeter-Shafer evidential theory.The effective and validity of the provided method is testified by the simulating result.更多还原